@avallon-labs/mcp 50.3.0 → 50.4.0

package/dist/index.js

@@ -46,5 +46,5 @@ function setupFetch() {
 
 // src/index.ts
 setupFetch();
-await import("./server-7RPXVFML.js");
+await import("./server-NVU3VZZX.js");
 //# sourceMappingURL=index.js.map

package/dist/{server-7RPXVFML.js → server-NVU3VZZX.js}

@@ -1418,12 +1418,20 @@ var createWorkerEvalFromRun = async (workerRunId, createWorkerEvalFromRunBody, o
 var getRunEvalUrl = (evalId) => {
   return `/v1/evals/${evalId}/runs`;
 };
-var runEval = async (evalId, options) => {
+var runEval = async (evalId, runEvalBody, options) => {
+  const getHeaders = (h) => {
+    if (!h) return {};
+    if (h instanceof Headers) return Object.fromEntries(h.entries());
+    if (Array.isArray(h)) return Object.fromEntries(h);
+    return h;
+  };
   const res = await fetch(
     getRunEvalUrl(evalId),
     {
       ...options,
-      method: "POST"
+      method: "POST",
+      headers: { "Content-Type": "application/json", ...getHeaders(options?.headers) },
+      body: JSON.stringify(runEvalBody)
     }
   );
   const body = [204, 205, 304].includes(res.status) ? null : await res.text();
@@ -4922,7 +4930,7 @@ var createWorkerEvalFromRunHandler = async (args, options) => {
   };
 };
 var runEvalHandler = async (args, options) => {
-  const res = await runEval(args.pathParams.evalId, options);
+  const res = await runEval(args.pathParams.evalId, args.bodyParams, options);
   if (res.status >= 400) {
     return {
       content: [
@@ -8807,17 +8815,9 @@ var CreateWorkerEvalFromRunParams = zod.object({
   "workerRunId": zod.string().uuid()
 });
 var createWorkerEvalFromRunBodyNameMax = 200;
-var createWorkerEvalFromRunBodyToolMocksItemDescriptionDefault = `Mocked tool available during an evaluation.`;
 var createWorkerEvalFromRunBodyGraderCriterionMax = 1e4;
 var CreateWorkerEvalFromRunBody = zod.object({
   "name": zod.string().min(1).max(createWorkerEvalFromRunBodyNameMax),
-  "tool_mocks": zod.array(zod.object({
-    "mcp_type": zod.string().min(1),
-    "tool_name": zod.string().min(1),
-    "description": zod.string().default(createWorkerEvalFromRunBodyToolMocksItemDescriptionDefault),
-    "input_schema": zod.record(zod.string(), zod.unknown()),
-    "responses": zod.array(zod.unknown()).min(1)
-  })).min(1),
   "grader": zod.object({
     "type": zod.literal("prompt"),
     "criterion": zod.string().min(1).max(createWorkerEvalFromRunBodyGraderCriterionMax)
@@ -8847,7 +8847,11 @@ var CreateWorkerEvalFromRunResponse = zod.object({
       "tool_name": zod.string(),
       "description": zod.string(),
       "input_schema": zod.record(zod.string(), zod.unknown()),
-      "responses": zod.array(zod.unknown())
+      "match": zod.record(zod.string(), zod.unknown()).optional(),
+      "logical_tool_name": zod.union([zod.string(), zod.null()]),
+      "responses": zod.array(zod.unknown()),
+      "inputs": zod.array(zod.unknown()).optional(),
+      "default": zod.unknown().optional()
     })),
     "grader": zod.object({
       "type": zod.literal("prompt"),
@@ -8860,6 +8864,10 @@ var CreateWorkerEvalFromRunResponse = zod.object({
 var RunEvalParams = zod.object({
   "evalId": zod.string().uuid()
 });
+var runEvalBodyWorkerVersionMax = 9007199254740991;
+var RunEvalBody = zod.object({
+  "worker_version": zod.number().int().min(1).max(runEvalBodyWorkerVersionMax)
+});
 var runEvalResponseDataTargetVersionMin = -9007199254740991;
 var runEvalResponseDataTargetVersionMax = 9007199254740991;
 var RunEvalResponse = zod.object({
@@ -8873,6 +8881,7 @@ var RunEvalResponse = zod.object({
     "error": zod.union([zod.string(), zod.null()]),
     "judge_model": zod.union([zod.string(), zod.null()]),
     "judge_prompt_version": zod.union([zod.string(), zod.null()]),
+    "tool_trace": zod.union([zod.unknown(), zod.null()]),
     "created_at": zod.string().datetime({ "offset": true }),
     "started_at": zod.union([zod.string().datetime({ "offset": true }), zod.null()]),
     "completed_at": zod.union([zod.string().datetime({ "offset": true }), zod.null()])
@@ -8895,6 +8904,7 @@ var GetEvalRunResponse = zod.object({
     "error": zod.union([zod.string(), zod.null()]),
     "judge_model": zod.union([zod.string(), zod.null()]),
     "judge_prompt_version": zod.union([zod.string(), zod.null()]),
+    "tool_trace": zod.union([zod.unknown(), zod.null()]),
     "created_at": zod.string().datetime({ "offset": true }),
     "started_at": zod.union([zod.string().datetime({ "offset": true }), zod.null()]),
     "completed_at": zod.union([zod.string().datetime({ "offset": true }), zod.null()])
@@ -12294,7 +12304,7 @@ var createMcpServer = (options) => {
     "createWorkerEvalFromRun",
     {
       title: "Create a worker eval from a run",
-      description: "Snapshot a worker run's trigger and resolved artifacts. Mock responses are returned in call order.",
+      description: "Snapshot a worker run's trigger and resolved artifacts, and record its MCP tool responses from telemetry for replay.",
       inputSchema: {
         pathParams: CreateWorkerEvalFromRunParams,
         bodyParams: CreateWorkerEvalFromRunBody.optional()
@@ -12311,9 +12321,10 @@ var createMcpServer = (options) => {
     "runEval",
     {
       title: "Run an eval",
-      description: "Queue the worker's current test version against the eval's immutable artifact snapshot and mocked tools.",
+      description: "Queue an explicit worker version against the eval's immutable artifact snapshot and mocked tools.",
       inputSchema: {
-        pathParams: RunEvalParams
+        pathParams: RunEvalParams,
+        bodyParams: RunEvalBody.optional()
       },
       outputSchema: RunEvalResponse,
       annotations: { destructiveHint: true }
@@ -13923,4 +13934,4 @@ var transport = new StdioServerTransport();
 server.connect(transport).then(() => {
   console.error("MCP server running on stdio");
 }).catch(console.error);
-//# sourceMappingURL=server-7RPXVFML.js.map
+//# sourceMappingURL=server-NVU3VZZX.js.map